Koichiro Yamaguchi's work experience spans multiple roles within various companies. Koichiro started in 2000 as a Researcher at NTT DOCOMO, Inc., where they focused on wireless network systems. From 2011 to 2013, they worked as a Visiting Researcher at the Toyota Technological Institute at Chicago, where they researched computer vision and developed dense matching methods.
In 2016, Yamaguchi joined the Toyota Research Institute as a Technical Advisor, working on computer vision research for autonomous driving. During their time there, they developed an efficient fully convolutional network (FCN) model for semantic segmentation and contributed to the development of the autonomous driving system.
From 2003 to 2019, Yamaguchi worked at Toyota Central R&D Labs., Inc. as a Senior Researcher. Koichiro focused on computer vision geometry and image recognition for advanced driver assistance systems. Koichiro also developed 3D recognition techniques and an automated generation system for high precision road maps.
In 2019, Yamaguchi moved to the Toyota Research Institute - Advanced Development as a Staff Engineer. There, they built an MLOps platform and developed a hyper-parameter optimization service to enhance the performance of ML models.
Currently, Yamaguchi is a Staff Engineer at Woven Planet Holdings, where they specialize in developing AutoML techniques to optimize ML models for target hardware. Koichiro also builds training frameworks and ML models for image recognition in various tasks.
Koichiro Yamaguchi has a strong educational background in the field of information science. Koichiro began their academic journey at Osaka University, where they obtained a Bachelor of Engineering degree in Information and Computer Science in 1998. Koichiro then pursued further specialization in the same field at the Nara Institute of Science and Technology, completing their Master of Engineering degree from 1998 to 2000. Building upon their expertise, they continued their education at the Nara Institute of Science and Technology and obtained a Doctor of Engineering degree in Information Science from 2006 to 2010.
January, 2021 - present